{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 10560,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.1915496587753296,
      "learning_rate": 4.75e-05,
      "loss": 0.7886,
      "step": 528
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.859733677829673,
      "eval_f1": 0.13302540415704386,
      "eval_loss": 0.4607163071632385,
      "eval_precision": 0.32432432432432434,
      "eval_recall": 0.08367228355607205,
      "eval_runtime": 4.523,
      "eval_samples_per_second": 206.723,
      "eval_steps_per_second": 3.316,
      "step": 528
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8614588975906372,
      "learning_rate": 4.5e-05,
      "loss": 0.3911,
      "step": 1056
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9292954387359634,
      "eval_f1": 0.6470908102229471,
      "eval_loss": 0.254240483045578,
      "eval_precision": 0.6080735820132857,
      "eval_recall": 0.6914584543869843,
      "eval_runtime": 4.6489,
      "eval_samples_per_second": 201.121,
      "eval_steps_per_second": 3.227,
      "step": 1056
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.2399094104766846,
      "learning_rate": 4.25e-05,
      "loss": 0.2384,
      "step": 1584
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9376428500447183,
      "eval_f1": 0.7163083377031987,
      "eval_loss": 0.19337689876556396,
      "eval_precision": 0.652651696129957,
      "eval_recall": 0.7937245787332946,
      "eval_runtime": 4.6125,
      "eval_samples_per_second": 202.712,
      "eval_steps_per_second": 3.252,
      "step": 1584
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.5993666648864746,
      "learning_rate": 4e-05,
      "loss": 0.1934,
      "step": 2112
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9445990261353473,
      "eval_f1": 0.7476784292915892,
      "eval_loss": 0.1678435504436493,
      "eval_precision": 0.68798828125,
      "eval_recall": 0.8187100522951772,
      "eval_runtime": 4.5469,
      "eval_samples_per_second": 205.634,
      "eval_steps_per_second": 3.299,
      "step": 2112
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.2465081214904785,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.172,
      "step": 2640
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9467852529066879,
      "eval_f1": 0.7566290364925177,
      "eval_loss": 0.1589244157075882,
      "eval_precision": 0.6901340996168582,
      "eval_recall": 0.8373038930854154,
      "eval_runtime": 4.5536,
      "eval_samples_per_second": 205.333,
      "eval_steps_per_second": 3.294,
      "step": 2640
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.3054755926132202,
      "learning_rate": 3.5e-05,
      "loss": 0.1602,
      "step": 3168
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9487727317897248,
      "eval_f1": 0.7631235309480282,
      "eval_loss": 0.15331855416297913,
      "eval_precision": 0.6930740037950665,
      "eval_recall": 0.8489250435793143,
      "eval_runtime": 4.6183,
      "eval_samples_per_second": 202.456,
      "eval_steps_per_second": 3.248,
      "step": 3168
    },
    {
      "epoch": 7.0,
      "grad_norm": 3.6129133701324463,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.1532,
      "step": 3696
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9497664712312431,
      "eval_f1": 0.7661898569570872,
      "eval_loss": 0.15049894154071808,
      "eval_precision": 0.693502824858757,
      "eval_recall": 0.8558977338756537,
      "eval_runtime": 4.5946,
      "eval_samples_per_second": 203.499,
      "eval_steps_per_second": 3.265,
      "step": 3696
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.5060432553291321,
      "learning_rate": 3e-05,
      "loss": 0.1457,
      "step": 4224
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9521514458908874,
      "eval_f1": 0.7754024808656638,
      "eval_loss": 0.14558807015419006,
      "eval_precision": 0.710348162475822,
      "eval_recall": 0.8535735037768739,
      "eval_runtime": 4.5124,
      "eval_samples_per_second": 207.206,
      "eval_steps_per_second": 3.324,
      "step": 4224
    },
    {
      "epoch": 9.0,
      "grad_norm": 1.1229195594787598,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.1401,
      "step": 4752
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9543376726622279,
      "eval_f1": 0.7870345566568443,
      "eval_loss": 0.1418333500623703,
      "eval_precision": 0.7301192842942346,
      "eval_recall": 0.8535735037768739,
      "eval_runtime": 4.662,
      "eval_samples_per_second": 200.56,
      "eval_steps_per_second": 3.218,
      "step": 4752
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.4011176824569702,
      "learning_rate": 2.5e-05,
      "loss": 0.1375,
      "step": 5280
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9550829772433668,
      "eval_f1": 0.7894174238375201,
      "eval_loss": 0.13877230882644653,
      "eval_precision": 0.7308263236021771,
      "eval_recall": 0.8582219639744335,
      "eval_runtime": 4.5718,
      "eval_samples_per_second": 204.514,
      "eval_steps_per_second": 3.281,
      "step": 5280
    },
    {
      "epoch": 11.0,
      "grad_norm": 1.9114675521850586,
      "learning_rate": 2.25e-05,
      "loss": 0.1331,
      "step": 5808
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.9554804730199742,
      "eval_f1": 0.7894174238375201,
      "eval_loss": 0.1359723061323166,
      "eval_precision": 0.7308263236021771,
      "eval_recall": 0.8582219639744335,
      "eval_runtime": 4.5381,
      "eval_samples_per_second": 206.032,
      "eval_steps_per_second": 3.305,
      "step": 5808
    },
    {
      "epoch": 12.0,
      "grad_norm": 1.272605061531067,
      "learning_rate": 2e-05,
      "loss": 0.1304,
      "step": 6336
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.954933916327139,
      "eval_f1": 0.784512683578104,
      "eval_loss": 0.13654367625713348,
      "eval_precision": 0.7257905138339921,
      "eval_recall": 0.8535735037768739,
      "eval_runtime": 4.5791,
      "eval_samples_per_second": 204.187,
      "eval_steps_per_second": 3.276,
      "step": 6336
    },
    {
      "epoch": 13.0,
      "grad_norm": 1.4920827150344849,
      "learning_rate": 1.75e-05,
      "loss": 0.1285,
      "step": 6864
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.9558779687965815,
      "eval_f1": 0.7906098219104155,
      "eval_loss": 0.13434641063213348,
      "eval_precision": 0.7380352644836272,
      "eval_recall": 0.8512492736780941,
      "eval_runtime": 4.6498,
      "eval_samples_per_second": 201.085,
      "eval_steps_per_second": 3.226,
      "step": 6864
    },
    {
      "epoch": 14.0,
      "grad_norm": 1.0049793720245361,
      "learning_rate": 1.5e-05,
      "loss": 0.1255,
      "step": 7392
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.9558779687965815,
      "eval_f1": 0.7958087049973133,
      "eval_loss": 0.13445836305618286,
      "eval_precision": 0.7401299350324838,
      "eval_recall": 0.8605461940732132,
      "eval_runtime": 4.5469,
      "eval_samples_per_second": 205.636,
      "eval_steps_per_second": 3.299,
      "step": 7392
    },
    {
      "epoch": 15.0,
      "grad_norm": 1.0454351902008057,
      "learning_rate": 1.25e-05,
      "loss": 0.1249,
      "step": 7920
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.954933916327139,
      "eval_f1": 0.7917669072440523,
      "eval_loss": 0.13459959626197815,
      "eval_precision": 0.7331683168316832,
      "eval_recall": 0.8605461940732132,
      "eval_runtime": 4.5061,
      "eval_samples_per_second": 207.495,
      "eval_steps_per_second": 3.329,
      "step": 7920
    },
    {
      "epoch": 16.0,
      "grad_norm": 1.3529750108718872,
      "learning_rate": 1e-05,
      "loss": 0.1238,
      "step": 8448
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.9551326642154427,
      "eval_f1": 0.7883328873427883,
      "eval_loss": 0.13417887687683105,
      "eval_precision": 0.7306547619047619,
      "eval_recall": 0.8558977338756537,
      "eval_runtime": 4.547,
      "eval_samples_per_second": 205.632,
      "eval_steps_per_second": 3.299,
      "step": 8448
    },
    {
      "epoch": 17.0,
      "grad_norm": 0.9557709693908691,
      "learning_rate": 7.5e-06,
      "loss": 0.1232,
      "step": 8976
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.9557289078803538,
      "eval_f1": 0.7904736419587904,
      "eval_loss": 0.13424266874790192,
      "eval_precision": 0.7326388888888888,
      "eval_recall": 0.8582219639744335,
      "eval_runtime": 4.6257,
      "eval_samples_per_second": 202.13,
      "eval_steps_per_second": 3.243,
      "step": 8976
    },
    {
      "epoch": 18.0,
      "grad_norm": 1.4014308452606201,
      "learning_rate": 5e-06,
      "loss": 0.1215,
      "step": 9504
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.954933916327139,
      "eval_f1": 0.7909212283044059,
      "eval_loss": 0.13506156206130981,
      "eval_precision": 0.7317193675889329,
      "eval_recall": 0.8605461940732132,
      "eval_runtime": 4.539,
      "eval_samples_per_second": 205.994,
      "eval_steps_per_second": 3.305,
      "step": 9504
    },
    {
      "epoch": 19.0,
      "grad_norm": 0.8569299578666687,
      "learning_rate": 2.5e-06,
      "loss": 0.1209,
      "step": 10032
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.9547351684388353,
      "eval_f1": 0.786648865153538,
      "eval_loss": 0.1337263584136963,
      "eval_precision": 0.7277667984189723,
      "eval_recall": 0.8558977338756537,
      "eval_runtime": 4.6671,
      "eval_samples_per_second": 200.337,
      "eval_steps_per_second": 3.214,
      "step": 10032
    },
    {
      "epoch": 20.0,
      "grad_norm": 1.4876703023910522,
      "learning_rate": 0.0,
      "loss": 0.1207,
      "step": 10560
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.95553015999205,
      "eval_f1": 0.7861921327267862,
      "eval_loss": 0.133191779255867,
      "eval_precision": 0.7286706349206349,
      "eval_recall": 0.8535735037768739,
      "eval_runtime": 4.6265,
      "eval_samples_per_second": 202.096,
      "eval_steps_per_second": 3.242,
      "step": 10560
    },
    {
      "epoch": 20.0,
      "step": 10560,
      "total_flos": 4541164131293502.0,
      "train_loss": 0.18862926418131049,
      "train_runtime": 1256.3234,
      "train_samples_per_second": 134.313,
      "train_steps_per_second": 8.405
    }
  ],
  "logging_steps": 500,
  "max_steps": 10560,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 4541164131293502.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}